{"id":"d2f12a9429c1a800f6085c0945218328720c8dda6c42ab509695f13bd71ca1cb","title":"METR reports forecasts on AI R&D acceleration and societal impacts","argument":"METR’s report, published August 20, 2025, presents a pilot survey by METR and the Forecasting Research Institute of AI forecasting domain experts and superforecasters. They estimated the chances of rapid AI progress following hypothetical AI research and development capabilities, and of extreme societal effects if progress accelerated. METR defines “effective compute” as a way to combine progress from larger compute spend and algorithmic improvements: the compute scale-up that would have been needed to match a more capable model without those innovations. The forecasters generally considered three-fold acceleration plausible but unlikely, while disagreeing sharply about possible extreme societal effects. METR says month-long, open-ended research evaluations could help resolve disagreement about acceleration, but the groups remained far apart on societal impacts even when forecasts assumed acceleration. METR concludes that this disagreement merits further study before expanding the pilot. It cautions that the small sample and subjective forecasts make the results suggestive, and that some aggregate statistics are fragile. The report does not establish whether the hypothetical capabilities or societal effects will occur.","topicIds":["artificial-general-intelligence"],"domain":"other","claimType":"attributed reading","limits":"Text as retrieved on October 23, 2026; originally published August 20, 2025. METR describes this as a pilot with a small sample and subjective judgment-based forecasts; it says some aggregate statistics are fragile. The report says some participants may have interpreted the forecasting concepts differently.","speaker":{"name":"METR","roleAtSource":"Publisher"},"source":{"title":"Forecasting the Impacts of AI R&D Acceleration: Results of a Pilot Study","publisher":"METR","format":"Article","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","publishedAt":"2025-08-20T07:00:00.000Z","dateBasis":"Date supplied by the publisher’s feed."},"evidence":[{"text":"METR and the Forecasting Research Institute conducted a pilot survey asking AI forecasting domain experts and superforecasters to estimate acceleration and societal-impact probabilities under hypothetical conditions.","excerpt":"we have worked with the Forecasting Research Institute to conduct an initial pilot survey of 8 AI forecasting domain experts1 (henceforth “experts”) and 10 “superforecasters”2","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"METR defines effective compute scale-up as the compute increase that would have been needed to match a more capable model using earlier algorithms, allowing time for adaptations.","excerpt":"“what factor more compute would have been needed to reach the performance of the more powerful model using algorithms available at the time when the earlier model was developed, if a team had ~3 months to make adaptations to the increased amount of compute?”","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"The forecasters considered three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with different median estimates from experts and superforecasters.","excerpt":"Experts and superforecasters both find a three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with experts giving it a median 20% chance and superforecasters 8%.","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"The groups disagreed strongly about the likelihood of extreme societal events; superforecasters’ median estimates were below 0.5% and experts’ were above 18%.","excerpt":"superforecasters giving a median risk estimate below 0.5% for all outcomes (compared to a median risk estimate of over 18% for all outcomes from the experts).","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"METR says the pilot’s small sample and reliance on subjective forecasts mean its results are suggestive, and that some aggregate statistics are fragile.","excerpt":"As this is a pilot study with only a small sample of participants, and relies on subjective judgment-based forecasts, these results should only be taken as suggestive evidence","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"}],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","checkedAt":null,"items":[]},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"What accounts for the sharp disagreement between experts and superforecasters about the societal effects of rapid AI progress?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"METR reports forecasts on AI R&D acceleration and societal impacts","summary":"METR’s report, published August 20, 2025, presents a pilot survey by METR and the Forecasting Research Institute of AI forecasting domain experts and superforecasters. They estimated the chances of rapid AI progress following hypothetical AI research and development capabilities, and of extreme societal effects if progress accelerated. METR defines “effective compute” as a way to combine progress from larger compute spend and algorithmic improvements: the compute scale-up that would have been needed to match a more capable model without those innovations. The forecasters generally considered three-fold acceleration plausible but unlikely, while disagreeing sharply about possible extreme societal effects. METR says month-long, open-ended research evaluations could help resolve disagreement about acceleration, but the groups remained far apart on societal impacts even when forecasts assumed acceleration. METR concludes that this disagreement merits further study before expanding the pilot. It cautions that the small sample and subjective forecasts make the results suggestive, and that some aggregate statistics are fragile. The report does not establish whether the hypothetical capabilities or societal effects will occur.","question":"What accounts for the sharp disagreement between experts and superforecasters about the societal effects of rapid AI progress?","claims":[{"text":"METR and the Forecasting Research Institute conducted a pilot survey asking AI forecasting domain experts and superforecasters to estimate acceleration and societal-impact probabilities under hypothetical conditions.","excerpt":"we have worked with the Forecasting Research Institute to conduct an initial pilot survey of 8 AI forecasting domain experts1 (henceforth “experts”) and 10 “superforecasters”2","support":"source-stated"},{"text":"METR defines effective compute scale-up as the compute increase that would have been needed to match a more capable model using earlier algorithms, allowing time for adaptations.","excerpt":"“what factor more compute would have been needed to reach the performance of the more powerful model using algorithms available at the time when the earlier model was developed, if a team had ~3 months to make adaptations to the increased amount of compute?”","support":"source-stated"},{"text":"The forecasters considered three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with different median estimates from experts and superforecasters.","excerpt":"Experts and superforecasters both find a three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with experts giving it a median 20% chance and superforecasters 8%.","support":"source-stated"},{"text":"The groups disagreed strongly about the likelihood of extreme societal events; superforecasters’ median estimates were below 0.5% and experts’ were above 18%.","excerpt":"superforecasters giving a median risk estimate below 0.5% for all outcomes (compared to a median risk estimate of over 18% for all outcomes from the experts).","support":"source-stated"},{"text":"METR says the pilot’s small sample and reliance on subjective forecasts mean its results are suggestive, and that some aggregate statistics are fragile.","excerpt":"As this is a pilot study with only a small sample of participants, and relies on subjective judgment-based forecasts, these results should only be taken as suggestive evidence","support":"source-stated"}],"limitations":["Text as retrieved on October 23, 2026; originally published August 20, 2025.","METR describes this as a pilot with a small sample and subjective judgment-based forecasts; it says some aggregate statistics are fragile.","The report says some participants may have interpreted the forecasting concepts differently."],"mediaId":null,"correctionReason":null,"id":"d2f12a9429c1a800f6085c0945218328720c8dda6c42ab509695f13bd71ca1cb","publicationId":"760eec9d-12e9-0eb2-3164-106be1f87901","version":1,"sourceUrl":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","sourceTitle":"Forecasting the Impacts of AI R&D Acceleration: Results of a Pilot Study","publisher":"METR","publishedAt":"2025-08-20T07:00:00.000Z","originalPublished":{"value":"2025-08-20T07:00:00.000Z","precision":"instant","basis":"backfill:feed_published"},"firstObservedAt":"2026-10-18T03:21:21.795Z","createdAt":"2026-11-12T12:23:23.243Z","historical":true,"entityIds":["metr"],"topicIds":["artificial-general-intelligence"],"media":[],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","checkedAt":null,"items":[]},"editorialDecision":{"outcome":"accept","mediaId":null,"inputHash":"d27688223f5cfb57beaa2bcb17a971b33ef5d18cd0eff13931a43b61d73a327a","rubricVersion":"unrecorded","provider":"vercel","model":"typesafe-ai/jev","reasonCodes":["bridge_admitted_by_model"],"confidence":0.71,"decidedAt":"2026-10-23T18:43:39.476Z"},"provenance":{"textHash":"1111351e03ee44299d807efd57f28a80b230d5dac5e3a40afe3fbffe33264026","observationId":"d6731c034bdb636d05b4a9a57bb79bb73b16438aaf644a3f616ffae797bd6922","retrievedAt":"2026-10-23T18:45:38.292Z","kind":"publisher-page","preparer":{"model":"openai/gpt-6-luna","runId":"9ffd4200-3e7a-d9d8-da3a-1e2d824f971f"},"reviewer":{"model":"anthropic/claude-opus-5.5","runId":"f3d0b3f9-602f-79f0-1cac-39f9b93b47c3"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-12T12:19:23.282Z","model":"anthropic/claude-opus-5.5"}},"canonicalUrl":"https://escapemeridian.com/readings/metr-reports-forecasts-on-ai-r-and-d-acceleration-and-societal-d2f12a94","formats":{"html":"/readings/metr-reports-forecasts-on-ai-r-and-d-acceleration-and-societal-d2f12a94","markdown":"/readings/metr-reports-forecasts-on-ai-r-and-d-acceleration-and-societal-d2f12a94.md","json":"/readings/metr-reports-forecasts-on-ai-r-and-d-acceleration-and-societal-d2f12a94.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[],"entities":[{"id":"metr","name":"METR","kind":"organization","context":"METR studies what AI systems can do and how reliably they complete tasks, including work with human developers.","referenceUrls":["https://metr.org/"],"path":"/entities/metr","json":"/entities/metr.json","markdown":"/entities/metr.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"d2f12a9429c1a800f6085c0945218328720c8dda6c42ab509695f13bd71ca1cb","revision":"760eec9d-12e9-0eb2-3164-106be1f87901","revisionKind":"publication-id","title":"METR reports forecasts on AI R&D acceleration and societal impacts","url":"https://escapemeridian.com/readings/metr-reports-forecasts-on-ai-r-and-d-acceleration-and-societal-d2f12a94"},"source":{"type":"source","id":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","revision":"1111351e03ee44299d807efd57f28a80b230d5dac5e3a40afe3fbffe33264026","revisionKind":"sha256-source-text","title":"Forecasting the Impacts of AI R&D Acceleration: Results of a Pilot Study","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"d2f12a9429c1a800f6085c0945218328720c8dda6c42ab509695f13bd71ca1cb","title":"METR reports forecasts on AI R&D acceleration and societal impacts","argument":"METR’s report, published August 20, 2025, presents a pilot survey by METR and the Forecasting Research Institute of AI forecasting domain experts and superforecasters. They estimated the chances of rapid AI progress following hypothetical AI research and development capabilities, and of extreme societal effects if progress accelerated. METR defines “effective compute” as a way to combine progress from larger compute spend and algorithmic improvements: the compute scale-up that would have been needed to match a more capable model without those innovations. The forecasters generally considered three-fold acceleration plausible but unlikely, while disagreeing sharply about possible extreme societal effects. METR says month-long, open-ended research evaluations could help resolve disagreement about acceleration, but the groups remained far apart on societal impacts even when forecasts assumed acceleration. METR concludes that this disagreement merits further study before expanding the pilot. It cautions that the small sample and subjective forecasts make the results suggestive, and that some aggregate statistics are fragile. The report does not establish whether the hypothetical capabilities or societal effects will occur.","topicIds":["artificial-general-intelligence"],"domain":"other","claimType":"attributed reading","limits":"Text as retrieved on October 23, 2026; originally published August 20, 2025. METR describes this as a pilot with a small sample and subjective judgment-based forecasts; it says some aggregate statistics are fragile. The report says some participants may have interpreted the forecasting concepts differently.","speaker":{"name":"METR","roleAtSource":"Publisher"},"source":{"title":"Forecasting the Impacts of AI R&D Acceleration: Results of a Pilot Study","publisher":"METR","format":"Article","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","publishedAt":"2025-08-20T07:00:00.000Z","dateBasis":"Date supplied by the publisher’s feed."},"evidence":[{"text":"METR and the Forecasting Research Institute conducted a pilot survey asking AI forecasting domain experts and superforecasters to estimate acceleration and societal-impact probabilities under hypothetical conditions.","excerpt":"we have worked with the Forecasting Research Institute to conduct an initial pilot survey of 8 AI forecasting domain experts1 (henceforth “experts”) and 10 “superforecasters”2","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"METR defines effective compute scale-up as the compute increase that would have been needed to match a more capable model using earlier algorithms, allowing time for adaptations.","excerpt":"“what factor more compute would have been needed to reach the performance of the more powerful model using algorithms available at the time when the earlier model was developed, if a team had ~3 months to make adaptations to the increased amount of compute?”","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"The forecasters considered three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with different median estimates from experts and superforecasters.","excerpt":"Experts and superforecasters both find a three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with experts giving it a median 20% chance and superforecasters 8%.","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"The groups disagreed strongly about the likelihood of extreme societal events; superforecasters’ median estimates were below 0.5% and experts’ were above 18%.","excerpt":"superforecasters giving a median risk estimate below 0.5% for all outcomes (compared to a median risk estimate of over 18% for all outcomes from the experts).","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"},{"text":"METR says the pilot’s small sample and reliance on subjective forecasts mean its results are suggestive, and that some aggregate statistics are fragile.","excerpt":"As this is a pilot study with only a small sample of participants, and relies on subjective judgment-based forecasts, these results should only be taken as suggestive evidence","url":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","relationship":"Source-stated"}],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","checkedAt":null,"items":[]},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"What accounts for the sharp disagreement between experts and superforecasters about the societal effects of rapid AI progress?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"METR reports forecasts on AI R&D acceleration and societal impacts","summary":"METR’s report, published August 20, 2025, presents a pilot survey by METR and the Forecasting Research Institute of AI forecasting domain experts and superforecasters. They estimated the chances of rapid AI progress following hypothetical AI research and development capabilities, and of extreme societal effects if progress accelerated. METR defines “effective compute” as a way to combine progress from larger compute spend and algorithmic improvements: the compute scale-up that would have been needed to match a more capable model without those innovations. The forecasters generally considered three-fold acceleration plausible but unlikely, while disagreeing sharply about possible extreme societal effects. METR says month-long, open-ended research evaluations could help resolve disagreement about acceleration, but the groups remained far apart on societal impacts even when forecasts assumed acceleration. METR concludes that this disagreement merits further study before expanding the pilot. It cautions that the small sample and subjective forecasts make the results suggestive, and that some aggregate statistics are fragile. The report does not establish whether the hypothetical capabilities or societal effects will occur.","question":"What accounts for the sharp disagreement between experts and superforecasters about the societal effects of rapid AI progress?","claims":[{"text":"METR and the Forecasting Research Institute conducted a pilot survey asking AI forecasting domain experts and superforecasters to estimate acceleration and societal-impact probabilities under hypothetical conditions.","excerpt":"we have worked with the Forecasting Research Institute to conduct an initial pilot survey of 8 AI forecasting domain experts1 (henceforth “experts”) and 10 “superforecasters”2","support":"source-stated"},{"text":"METR defines effective compute scale-up as the compute increase that would have been needed to match a more capable model using earlier algorithms, allowing time for adaptations.","excerpt":"“what factor more compute would have been needed to reach the performance of the more powerful model using algorithms available at the time when the earlier model was developed, if a team had ~3 months to make adaptations to the increased amount of compute?”","support":"source-stated"},{"text":"The forecasters considered three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with different median estimates from experts and superforecasters.","excerpt":"Experts and superforecasters both find a three-fold acceleration of effective compute scale-up by 2029 plausible but unlikely, with experts giving it a median 20% chance and superforecasters 8%.","support":"source-stated"},{"text":"The groups disagreed strongly about the likelihood of extreme societal events; superforecasters’ median estimates were below 0.5% and experts’ were above 18%.","excerpt":"superforecasters giving a median risk estimate below 0.5% for all outcomes (compared to a median risk estimate of over 18% for all outcomes from the experts).","support":"source-stated"},{"text":"METR says the pilot’s small sample and reliance on subjective forecasts mean its results are suggestive, and that some aggregate statistics are fragile.","excerpt":"As this is a pilot study with only a small sample of participants, and relies on subjective judgment-based forecasts, these results should only be taken as suggestive evidence","support":"source-stated"}],"limitations":["Text as retrieved on October 23, 2026; originally published August 20, 2025.","METR describes this as a pilot with a small sample and subjective judgment-based forecasts; it says some aggregate statistics are fragile.","The report says some participants may have interpreted the forecasting concepts differently."],"mediaId":null,"correctionReason":null,"id":"d2f12a9429c1a800f6085c0945218328720c8dda6c42ab509695f13bd71ca1cb","publicationId":"760eec9d-12e9-0eb2-3164-106be1f87901","version":1,"sourceUrl":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","sourceTitle":"Forecasting the Impacts of AI R&D Acceleration: Results of a Pilot Study","publisher":"METR","publishedAt":"2025-08-20T07:00:00.000Z","originalPublished":{"value":"2025-08-20T07:00:00.000Z","precision":"instant","basis":"backfill:feed_published"},"firstObservedAt":"2026-10-18T03:21:21.795Z","createdAt":"2026-11-12T12:23:23.243Z","historical":true,"entityIds":["metr"],"topicIds":["artificial-general-intelligence"],"media":[],"sourceMedia":{"status":"unchecked","sourceUrl":"https://metr.org/blog/2025-08-20-forecasting-impacts-of-ai-acceleration/","checkedAt":null,"items":[]},"editorialDecision":{"outcome":"accept","mediaId":null,"inputHash":"d27688223f5cfb57beaa2bcb17a971b33ef5d18cd0eff13931a43b61d73a327a","rubricVersion":"unrecorded","provider":"vercel","model":"typesafe-ai/jev","reasonCodes":["bridge_admitted_by_model"],"confidence":0.71,"decidedAt":"2026-10-23T18:43:39.476Z"},"provenance":{"textHash":"1111351e03ee44299d807efd57f28a80b230d5dac5e3a40afe3fbffe33264026","observationId":"d6731c034bdb636d05b4a9a57bb79bb73b16438aaf644a3f616ffae797bd6922","retrievedAt":"2026-10-23T18:45:38.292Z","kind":"publisher-page","preparer":{"model":"openai/gpt-6-luna","runId":"9ffd4200-3e7a-d9d8-da3a-1e2d824f971f"},"reviewer":{"model":"anthropic/claude-opus-5.5","runId":"f3d0b3f9-602f-79f0-1cac-39f9b93b47c3"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-12T12:19:23.282Z","model":"anthropic/claude-opus-5.5"}}}],"audits":[{"publicationId":"760eec9d-12e9-0eb2-3164-106be1f87901","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading attributes the pilot survey, the effective-compute definition and the conclusions to METR, working with the Forecasting Research Institute. This matches the single METR blog post in the evidence. Forecasts are kept as participants' estimates rather than presented as facts.","excerpt":"we have worked with the Forecasting Research Institute to conduct an initial pilot survey"},{"dimension":"summary","status":"uncertain","explanation":"Two central terms are left unexplained, and the definitions list is empty. First, 'superforecasters' is never defined; the source footnote calls them top performers from forecasting tournaments or platforms. Second, the summary says forecasters found 'three-fold acceleration' plausible but unlikely without saying what is accelerating. In the source, it means effective compute scale-up, with a year of 2018–2024 progress happening every 4 months. Smallest repair: add a short gloss for superforecasters and name the three-fold acceleration of effective compute scale-up in plain words. The summary also omits the medians (20% vs 8%), though the claims include them.","excerpt":"the amount of progress that happened in one year between 2018 and 2024 now happens every 4 months"},{"dimension":"date","status":"pass","explanation":"evidence.originalPublished is the instant 2025-08-20T07:00Z, whose UTC calendar day is August 20, 2025; this matches the URL path and the reading's stated date. The reading presents forecasts as forecasts, not outcomes, and uses no hindsight framing. The retrieval date appears only as a capture note.","excerpt":null},{"dimension":"support","status":"pass","explanation":"Each claim matches its quoted source text: the survey design, the effective-compute definition, the 20%/8% medians, the below-0.5% vs over-18% split, and the pilot caveats. The summary says month-long research evaluations could resolve the acceleration disagreement but not the societal gap; this is supported by the conclusion and by the conditional results. No benefits or costs are invented. The closing question follows from the source's own call to explore the disagreement.","excerpt":"Conditioning on a three-fold (or even ten-fold) acceleration does not close this gap in the predictions of the two groups"},{"dimension":"media","status":"pass","explanation":"No media is selected, and the evidence supplies no verified media candidates. No link has been invented.","excerpt":null}],"createdAt":"2026-11-12T12:19:23.282Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}